Where the Time Actually Goes
The average small-molecule IND-to-approval timeline is 12 years. The average large-molecule is longer. Almost everyone in biotech accepts this as the cost of doing business. It shouldn't be.
When you actually decompose the timeline — milestone by milestone, function by function — the picture that emerges is not what most people expect. The science itself, the actual experiments and trials, accounts for maybe 40% of elapsed time. The remaining 60% lives in four categories:
1. Sequential dependencies where parallel execution was possible. A regulatory team waits for clinical data that was available three months ago but wasn't surfaced. A CMC section waits for a formatting review that takes two weeks for work that takes two hours.
2. Document creation and review cycles. An NDA module goes through five rounds of internal review before it reaches the agency. Each round adds three to six weeks. The substantive changes between round three and round five are typically minor.
3. Coordination and communication overhead. A cross-functional decision that requires input from five people takes four to six weeks. Not because any individual needs that long, but because calendars don't align, emails pile up, and context gets lost in handoffs.
4. Search, synthesis, and analysis tasks. A regulatory strategist spends 40% of their week reading and synthesizing — literature, guidance documents, precedent decisions. A bench scientist spends comparable time on literature review. This work is necessary, but it doesn't have to take as long as it does.
The compression question
The right question isn't "how do we make everyone work faster." That's not sustainable and it's not where the leverage is. The right question is: which of these four categories can AI compress without compromising the quality of the output?
Categories 1 and 3 are coordination problems — they require organizational redesign, not just better tools. Categories 2 and 4 are where AI creates immediate, measurable leverage. Document creation that takes days can be done in hours. Literature synthesis that takes a week can be done in an afternoon.
Over the next four modules, you'll build a precise map of where your organization's time actually lives — and calculate the compression available to you right now, with tools that exist today.
Exercise: Your timeline audit
Before the next lesson, sketch your current critical path for your most important program. Mark each step with one of four labels: Science (can't compress without compromising data), Sequential (could be parallel), Documents (AI-compressible), or Coordination (org design problem). Most people find that 50-60% of their timeline is in the last three categories.
That's your opportunity.